- Haskell-Cafe - Haskell.org

Re: [Haskell-cafe] ANN: io-memoize 1.1
by Peter Simons 13 Jun '14

13 Jun '14

Hi Dan, > 1. Create an empty cache with newCache > 2. fetch repeatedly from the cache, each time providing a fallback in > case it is empty. > > Fetching from a Cache is thread safe: only one "fallback" will execute > at a time, and after the first successful fallback, the cached value > will be set and no other fallbacks will be called. I'm not sure whether I get this correctly. If multiple threads fetch values from one cache, will only one of those threads run at a time? Or will those that fetch pairwise different values still run concurrently? In other words, does the cache have a global lock or is there a lock per value? Best regards, Peter

2 1

Not in scope
by Julian K. Arni 13 Jun '14

13 Jun '14

I have a lot of boilerplate code that th-desugar has made significantly more manageable (thanks Richard!). Somewhere in the midst of it, I have: >> type_equal (DContT n1) (DContT n2) = {...} Which, when I compile with cabal, gives me: src/Language/Haskell/TH/Alpha.hs:180:13: Not in scope: data constructor ‘DContT’ Perhaps you meant one of these: ‘DConT’ (imported from Language.Haskell.TH.Desugar), ‘DConE’ (imported from Language.Haskell.TH.Desugar), ‘DConPa’ (imported from Language.Haskell.TH.Desugar) src/Language/Haskell/TH/Alpha.hs:180:25: Not in scope: data constructor ‘DContT’ Perhaps you meant one of these: ‘DConT’ (imported from Language.Haskell.TH.Desugar), ‘DConE’ (imported from Language.Haskell.TH.Desugar), ‘DConPa’ (imported from Language.Haskell.TH.Desugar) Huh? Of course 'DConT' is what I meant - that's exactly what I wrote! What in the world is going on? GHC is happy with all the other constructors! I tried rewriting, just to be sure I didn't *somehow* get a tricky unicode character in there...

3 2

Cross compiling Haskell from Linux to Windows with GHCs LLVM backend?
by Rob Stewart 13 Jun '14

13 Jun '14

Hi, I compile my code with GHC on Linux, and don't have ready access to a Windows or a Mac operating system. Is it possible to use the GHC LLVM backend as a means to cross compile Haskell code to target multiple operating systems? I had envisaged it possible to generate the LLVM IR using GHCs -keep-llvm-files flag, and using LLVMs target triple flag to specify the CPU architecture and operating system of the users machine. I'd like to compile on Linux, to target WIndows and Mac machines. Strangely I cannot find blog posts or tutorials online for cross-compiling Haskell with GHC via LLVM. Is it possible? Thanks, -- Rob

4 4

Re: [Haskell-cafe] [Haskell] ANN: io-memoize 1.1
by Kim-Ee Yeoh 13 Jun '14

13 Jun '14

On Thu, Jun 12, 2014 at 4:47 PM, Dan Burton <danburton.email(a)gmail.com> wrote: > Have you ever wanted to restrict an IO action such that it could only be > invoked once? Now you can with System.IO.Memoize.once (formerly known as > ioMemo). > (Moving to cafe for dicussion ...) I may want a one-time only IO action, although so far the need hasn't arisen yet. ;) In the spirit of pilfering while the buns are still hot out of the oven, is there a backstory for this package you could possibly sketch out? Some canonical use case maybe? -- Kim-Ee

4 4

Opening for OpenGL, C++ Development.
by Manoj Chaudhari 12 Jun '14

12 Jun '14

Hello, We are looking for Software Developers with OpenGL experience. - Should have experience in C++/Xcode, OpenGL/DirectX/IOS. - Experience with Shafer programming, Multi-threaded applications, Graphics in CAD domain. - The role is of Software Development on primary CAD application like CATIA. - Good skills in Geometry, algorithms, Mathematics, API's, Graphics-will be a added advantage. Experience : 2 to 6 years. If anyone is interested on learning more about this opportunity and potential salary ranges, please email me at : manoj_chaudhari2(a)3dplmsoftware.com Regards, Manoj Chaudhari

3 2

Troubles installing EclipseFP
by Vagif Verdi 12 Jun '14

12 Jun '14

Hi, i'm trying to install eclipsefp. Ubuntu 14.04, ghc 7.6.3, cabal-install 1.20. I did it with a fresh .cabal and .ghc folders. Nothing else was installed. After restarting eclipse it started doenloading and building haskell libs, and stopped with this error: Cabal==1.16.0/installed-c6e..., SourceGraph => Cabal<1.7) rejecting: SourceGraph-0.2, 0.1 (conflict: buildwrapper => Cabal==1.16.0/installed-c6e..., SourceGraph => Cabal<1.5) Backjump limit reached (change with --max-backjumps). Note: when using a sandbox, all packages are required to have consistent dependencies. Try reinstalling/unregistering the offending packages or recreating the sandbox.

1 0

ghc-options: -dynamic in package refused by Hackage
by JP Moresmau 12 Jun '14

12 Jun '14

Hello, what's the recommended way to make a package dynamic? 1. I need to make the executable in my package dynamic otherwise I run into https://ghc.haskell.org/trac/ghc/ticket/8376 2. I've set ghc-options: -dynamic in my Cabal file and that solves my issues. 3. I can't upload my package to Hackage, I get a "'ghc-options: -d*' debug flags are not appropriate for a distributed package" error. I hadn't realized -dynamic is for debugging purposes... So is there another way to make my package dynamic and uploadable to Hackage? Or should I just not put the flag in the cabal flag and let my users pass the flag when they install? Thanks! JP -- JP Moresmau http://jpmoresmau.blogspot.com/

1 0

Simple but slow, and complex but fast Forth emulation [Was: Bang, a drum DSL for Haskell]
by oleg＠okmij.org 12 Jun '14

12 Jun '14

Emulation of a simple subset of Forth in Haskell seems easy. The trick, continuation-passing style, has been known for a long time. The trick underlies `functional unparsing' by Olivier Danvy. http://www.brics.dk/RS/98/12/BRICS-RS-98-12.pdf (published in JFP in 1998). Chris Okasaki later extended the technique http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.121.1890 http://dl.acm.org/citation.cfm?id=581699 He also noted huge types and slow compilation times. But there is another way. It is far more complex, and far fast. It is used in HSXML, which handles polyvariadic functions with literally thousands of arguments (some of my web pages are long). The following complete code illustrates the idea. {-# LANGUAGE MultiParamTypeClasses #-} {-# LANGUAGE FlexibleInstances, FlexibleContexts #-} {-# LANGUAGE TypeFamilies #-} -- Simple Forth module SimpleForth where -- The simple way: Danvy's Functional unparsing begin k = k () n st x k = k ((x::Int),st) add (n1,(n2,st)) k = k (n1+n2,st) end (top,st) = top t1 = begin n 1 n 2 add end -- Uncomment the following to get a hint why t2 is too slow -- t1 = begin _h n 1 n 2 add end -- Uncomment the following only if you are prepared to wait {- t2 = begin n 1 n 2 n 3 n 4 n 5 n 6 n 7 n 8 n 9 add add add add add add add add end -} -- A more complex but faster way -- Start with a stack 'stack' and then continue class Forth stack r where build :: stack -> r data End = End instance (a ~ stack) => Forth stack (End -> a) where build stack _ = stack data Add = Add -- Start with (Int, (Int, stack)), see Add and continue with (Int,stack) instance Forth (Int,stack) r => Forth (Int,(Int,stack)) (Add -> r) where build (n1,(n2,stack)) _ = build (n1+n2,stack) data N = N instance (a ~ Int, Forth (Int,stack) r) => Forth stack (N -> a -> r) where build stack _ n = build (n,stack) -- All of the following typecheck instantaneously, even on my slow -- laptop tt1 = build () N 1 N 2 Add End tt2 = build () N 1 N 2 N 3 N 4 N 5 N 6 N 7 N 8 N 9 Add Add Add Add Add Add Add Add End tt3 = build () N 1 N 2 Add N 3 Add N 4 Add N 5 Add N 6 Add N 7 Add N 8 Add N 9 Add End

1 0

Deleting mailing list
by Jürgen Nicklisch-Franken 12 Jun '14

12 Jun '14

I want to delete the mailman mailing list for leksah (leksah(a)projects.haskell.org.) since it is not used anymore and only collects spam. It looks as I can't do this on my own, so I need help. My mails to support(a)community.haskell.org didn't produce any response. So I try this for years now. Please! Is someone out there that can help me? Kind regards Jürgen Nicklisch-Franken

1 0

Unable to make a HList-like to deal with Kleisli Arrows and State
by Gautier DI FOLCO 12 Jun '14

12 Jun '14

Hi all, My goal is to compose Kleisli arrows with (State s) as parameteric Monadic type. So, I end up with the following code: data REnv :: [*] -> * -> * -> (* -> * -> *) -> ((* -> *) -> * -> * -> *) -> * where Last :: k (s t) i o -> REnv (t ': '[]) i o s k RCons :: k (s t) i o' -> REnv ts o' o s k -> REnv (t ': ts) i o s k Which, of course, gives me the following error when I try to play with it: *H> :t Last genRanking <interactive>:1:6: Kind incompatibility when matching types: s :: * -> * -> * StateT Ranking :: (* -> *) -> * -> * Expected type: Kleisli (s t) Ticket Ranking Actual type: Kleisli (State Ranking) Ticket Ranking In the first argument of ‘Last’, namely ‘genRanking’ In the expression: Last genRanking <interactive>:1:6: Kind incompatibility when matching types: t :: * Data.Functor.Identity.Identity :: * -> * Expected type: Kleisli (s t) Ticket Ranking Actual type: Kleisli (State Ranking) Ticket Ranking In the first argument of ‘Last’, namely ‘genRanking’ In the expression: Last genRanking*H> :t LastLast :: k (s t) i o -> REnv '[t] i o s k*H> :t genRanking genRanking :: Kleisli (State Ranking) Ticket Ranking I understand why it fails (State is * -> * -> *, while s is * -> *), but I can't figure out how to write this, I tried (s t o), but it isn't correct because k waits for a (* -> *). If you have any ideas, I'm curious to read them. Thanks in advance for your help.

2 2